@hdwebsoft/hdcode-ai-darwin-x64 0.0.7 → 0.0.8

package/resources/skills/hd-daily-viewer/SKILL.md

@@ -0,0 +1,222 @@
+---
+name: hd-daily-viewer
+description: "View, navigate, and summarize past daily goals and reports saved to ~/.hd/. Supports single-day detail view, time-window summaries with aggregated metrics, and a list mode showing recent file availability. Use when reviewing past work, comparing goals vs results, or getting a weekly/monthly overview."
+license: proprietary
+metadata:
+  version: "1.0.0"
+  copyright: "© HDWEBSOFT. All rights reserved."
+---
+
+# Daily Viewer Skill
+
+> Read-only. Never writes or modifies files. Data sources: `~/.hd/daily-goals/` and `~/.hd/daily-reports/`.
+
+---
+
+## Startup
+
+> This skill never writes files. Config is read-only. Source files in `~/.hd/` are never modified.
+
+Read `~/.hd/config.yaml` if it exists → load `ticket_patterns` into `$URL_PATTERN[PREFIX]`. If config is absent or empty: proceed normally with no linkification. No history scan. No config write under any circumstances.
+
+---
+
+## Mode Detection
+
+Parse invocation args before doing anything else:
+
+| Input | Mode |
+|-------|------|
+| No args | **List mode** |
+| `today`, `yesterday`, `last Monday`, `YYYY-MM-DD`, `March 3`, any single date reference | **Single-day mode** |
+| `this week`, `last week`, `last N days`, `last month`, `March`, `YYYY-MM-DD to YYYY-MM-DD`, any multi-day range | **Window mode** |
+
+Resolve relative dates using today's date. "Last Monday" = the most recent Monday before today.
+
+---
+
+## Data Sources
+
+```bash
+ls ~/.hd/daily-goals/    # files: YYYY-MM-DD.md
+ls ~/.hd/daily-reports/  # files: YYYY-MM-DD.md
+```
+
+For any given date, either or both files may exist. Always handle missing gracefully:
+- No goals file → show `[no goals recorded]`
+- No report file → show `[no report recorded]`
+- Neither file → show `[no data for this date]`
+
+---
+
+## List Mode (no args)
+
+Show the last 7 calendar days (regardless of weekends) with availability indicators.
+
+```
+📅 Recent daily files:
+
+  2026-03-04  ✅ goals  ✅ report     ← today
+  2026-03-03  ✅ goals  ✅ report
+  2026-03-02  ✅ goals  ❌ report
+  2026-03-01  ❌ goals  ✅ report
+  2026-02-28  ✅ goals  ✅ report
+  2026-02-27  ❌ goals  ❌ report
+  2026-02-26  ✅ goals  ✅ report
+
+What would you like to view?
+  • A specific day: "yesterday", "March 3", "2026-02-28"
+  • A time window:  "this week", "last week", "last 7 days", "February"
+```
+
+Wait for user input, then switch to single-day or window mode accordingly.
+
+---
+
+## Single-Day Mode
+
+Show full detail — goals and report side by side in clearly separated sections.
+
+```
+📅 {WEEKDAY}, {MONTH} {DAY}, {YEAR}
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+🎯 GOALS
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+{full content of ~/.hd/daily-goals/YYYY-MM-DD.md}
+[no goals recorded]  ← if file missing
+
+📋 REPORT
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+{full content of ~/.hd/daily-reports/YYYY-MM-DD.md}
+[no report recorded]  ← if file missing
+
+📊 DAY METRICS
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+Goals set:    {N} across {P} projects
+Done tasks:   {N}
+In-progress:  {N}
+Blockers:     {N} total → {X} solved ✅  {Y} still open ⚠️
+```
+
+**Linkification pass:** Before rendering any file content, scan each line for bare ticket IDs matching a known prefix in `$URL_PATTERN` (pattern `[A-Z]+-\d+` where prefix is in `$URL_PATTERN` and value ≠ `"no-tracker"`). Replace bare `PREFIX-NNNN` with `[PREFIX-NNNN](expanded_url)` in display output only. IDs already formatted as markdown links are left unchanged. IDs with no known pattern render as-is. Source files are never modified.
+
+Parse metrics directly from the file contents. If a file is missing, show `—` for its metrics.
+
+After display, ask:
+```
+View another day or time window? (or press Enter to exit)
+```
+
+---
+
+## Window Mode
+
+### Step 1 — Resolve window to a date range
+
+Compute start and end dates:
+- `this week` → Monday of current week to today
+- `last week` → Monday to Friday of the previous week
+- `last 7 days` → today minus 6 days to today
+- `last 30 days` / `last month` → today minus 29 days to today
+- `March` / `February` → full calendar month (current year assumed)
+- `YYYY-MM-DD to YYYY-MM-DD` → exact range
+
+### Step 2 — Collect data for all dates in range
+
+For each date: read `~/.hd/daily-goals/YYYY-MM-DD.md` and `~/.hd/daily-reports/YYYY-MM-DD.md` if they exist.
+
+### Step 3 — Display (detail level scales with window size)
+
+| Window size | Detail level |
+|-------------|-------------|
+| 1 day | → use Single-Day Mode instead |
+| 2–3 days | Per-day breakdown with task-level lists |
+| 4–7 days | Per-day summary row + aggregated metrics |
+| > 7 days | High-level aggregates + weekly rows, no per-day task lists |
+
+**Regardless of window size: always show at least 7 Outstanding Activities** (see below).
+
+---
+
+### Summary Header (all window sizes)
+
+```
+📊 Summary: {START_DATE} – {END_DATE} ({N} days, {D} days with data)
+
+✅ Done tasks:        {N}
+⚠️ Blockers:          {N} total → {X} solved ✅  {Y} still open ⚠️
+🎯 Goals set:         {N} → {A} achieved, {P} partial, {M} missed
+⏱️ Time-consuming tasks (top 3 by total hours logged across window):
+   - [{TASK-ID}] {Task name} — {Nh} across {D} days
+   - [{TASK-ID}] {Task name} — {Nh}
+   - [{TASK-ID}] {Task name} — {Nh}
+```
+
+Derive "time-consuming" from TIME ALLOCATION sections in daily reports — tasks or projects that appear across multiple days or with the highest single-day hours.
+
+---
+
+### Per-Day Breakdown (window ≤ 7 days)
+
+```
+Per-day breakdown:
+  {WEEKDAY} {DATE} │ Goals: {N} │ Done: {N} │ In-progress: {N} │ Blockers: {summary}
+  {WEEKDAY} {DATE} │ Goals: {N} │ Done: {N} │ In-progress: {N} │ No report
+  ...
+```
+
+For window ≤ 3 days: also expand each row to show task-level lists under it (same format as single-day report sections, condensed). Apply the same linkification pass (bare ticket IDs → markdown links) to all task text rendered here.
+
+---
+
+### Long Window Aggregates (window > 7 days)
+
+```
+Weekly breakdown:
+  Week of {DATE}  │ {N} done │ {N} blockers ({X} solved) │ {N} goals set
+  Week of {DATE}  │ {N} done │ {N} blockers ({X} solved) │ {N} goals set
+  ...
+
+Top projects by time:
+  {Project A}: {N}h  │  {Project B}: {N}h  │  {Project C}: {N}h
+
+Recurring blockers (appeared on 2+ days):
+  - "{Blocker description}" — appeared {N} days
+```
+
+---
+
+### Outstanding Activities (always shown, minimum 7)
+
+Picked from: tasks that are IN PROGRESS or have open (unsolved) blockers within the window, sorted by most recent occurrence first.
+
+**If the window contains fewer than 7 such items:** expand backward day-by-day beyond the window start until 7 items are found. Note when items come from outside the window:
+
+```
+🔥 Outstanding Activities ({N} items):
+  1. [{TASK-ID}] {Task name} — {completion%} — due {when} ⚠️ open blocker: "{blocker text}"
+  2. [{TASK-ID}] {Task name} — {completion%} — due {when}
+  3. [{TASK-ID}] {Task name} — blocked: "{blocker text}"
+  ...
+  7. [{TASK-ID}] {Task name} — {completion%}   ← from {DATE}, outside window
+```
+
+Apply the linkification pass to all task IDs in the Outstanding Activities list.
+
+Parse these from:
+- `🔄 IN PROGRESS` sections in daily reports (completion %, when complete)
+- `⚠️ ISSUES & BLOCKERS` sections in daily reports (severity, status, note)
+
+Tasks with status `solved` in blockers are NOT outstanding. Only `open` or `in discussion`.
+
+---
+
+## After Display
+
+Always end with:
+```
+View something else? (specific date, "this week", "last month", or press Enter to exit)
+```
+
+Accept any date/range input and switch mode accordingly. Loop until user exits.

package/resources/skills/hd-docs-init/SKILL.md

@@ -244,6 +244,7 @@ Based on analysis, determine which docs to create using doc-mapping conventions.
 | Migration       | `docs/MIGRATION.md`                       |
 | Security standards  | `docs/SECURITY_STANDARDS.md` |
 | Coding standards    | `docs/CODING_STANDARDS.md`   |
+| Review standards    | `docs/REVIEW_STANDARDS.md`   |
 
 > **AGENTS.md vs CLAUDE.md**: `AGENTS.md` is the full development guidelines file read by Amp agents. `CLAUDE.md` is a short Claude Code configuration file that points Claude agents to `AGENTS.md`. Always generate both — they serve different agent runtimes but the same guidelines.
 
@@ -259,6 +260,8 @@ Task() → docs/ARCHITECTURE.md generator
 Task() → Package-specific AGENTS.md generators
 Task() → docs/SECURITY_STANDARDS.md generator (if any security signals detected in scan)
 Task() → docs/CODING_STANDARDS.md generator (always)
+Task() → docs/REVIEW_STANDARDS.md generator (always)
+Task() → docs/KNOWN_ISSUES.md generator (always)
 ```
 
 ### 3.3 Doc Templates
@@ -437,6 +440,32 @@ Always generate `docs/CODING_STANDARDS.md` on every hd-docs-init run (no signal
 
 4. Write to `docs/CODING_STANDARDS.md`
 
+---
+
+### 3.6 REVIEW_STANDARDS.md Generation
+
+Always generate `docs/REVIEW_STANDARDS.md` on every hd-docs-init run (no signal required).
+
+1. Start from the base template at `skills/hd-code-review/REVIEW_STANDARDS.md`
+
+2. Map detected project type to `tech_stack`:
+   | Project type (from Phase 1.2) | tech_stack value |
+   |-------------------------------|-----------------|
+   | Node.js / TypeScript          | nodejs          |
+   | Node.js Monorepo              | nodejs          |
+   | .NET Core / .NET 5+           | dotnet          |
+   | .NET Framework                | dotnet          |
+   | .NET Solution (multi-project) | dotnet          |
+   | Mixed (e.g. .NET API + Node frontend) | [dotnet, nodejs] |
+   | Other / unknown               | ~ (leave blank) |
+
+3. Pre-fill `tech_stack:` with the detected value as a convenience hint. The skill also auto-detects from diff
+   extensions at review time, so this field can remain `~` if left unchanged.
+   Add comment at top: `# Generated by docs-init — tech_stack is optional (auto-detected from diff); set only to override`
+   Add reference: `# Base template: see hd-code-review/REVIEW_STANDARDS.md in your Claude skills directory (~/.claude/skills/hd-code-review/REVIEW_STANDARDS.md)`
+
+4. Write to `docs/REVIEW_STANDARDS.md`
+
 ## Phase 4: Review
 
 ### 4.1 Validation via finder (Amp) / Explore Subagent (Claude)
@@ -553,6 +582,8 @@ User: "Initialize docs for this new project"
 - [ ] AGENTS.md generated (always required)
 - [ ] CLAUDE.md generated pointing to AGENTS.md (always required)
 - [ ] docs/CODING_STANDARDS.md generated (always required)
+- [ ] docs/REVIEW_STANDARDS.md generated (always required)
+- [ ] docs/KNOWN_ISSUES.md generated (always required)
 ```
 
 ## Output Summary Notes
@@ -566,6 +597,8 @@ After generation, report the following to the user:
 | `docs/ARCHITECTURE.md` | created | Verify diagrams match code |
 | `docs/SECURITY_STANDARDS.md` | created | Set `applicable_compliance` if security signals were detected |
 | `docs/CODING_STANDARDS.md` | created | Edit `docs/CODING_STANDARDS.md` to enable project policies (feature flags, observability, i18n) as `required: yes` |
+| `docs/REVIEW_STANDARDS.md` | created | Verify `tech_stack` is correct; add `custom_aspects` and `aspect_escalations` as needed |
+| `docs/KNOWN_ISSUES.md` | created | Populate with project-specific known issues and accepted configuration debt |
 
 ## Troubleshooting
 

package/resources/skills/hd-docs-parse/SKILL.md

@@ -51,6 +51,7 @@ Parse the invocation:
 | `--metadata` | No | off | Include document metadata in output |
 | `--json` | No | — | Output as JSON (content + metadata) |
 | `--check` | No | — | Verify installation and dependencies |
+| `--cache` | No | off | Enable result caching (speeds up repeated extractions) |
 
 ### Natural Language Resolution
 
@@ -278,4 +279,5 @@ The script prints clear error messages for common issues:
 /hd-docs-parse report.pdf --force-ocr             — force OCR on text PDF
 /hd-docs-parse report.pdf --metadata --json       — full metadata as JSON
 /hd-docs-parse --check                            — verify dependencies
+/hd-docs-parse report.pdf --cache                — enable caching for repeated extractions
 ```

package/resources/skills/hd-docs-parse/scripts/parse_document.py

@@ -78,6 +78,11 @@ def build_config(args):
 
     kwargs = {}
 
+    # Caching off by default; use --cache to enable (pins dir so hits work cross-session)
+    kwargs["use_cache"] = getattr(args, "cache", False)
+    if kwargs["use_cache"]:
+        kwargs["cache_dir"] = str(Path.home() / ".cache" / "kreuzberg")
+
     # OCR
     if args.ocr or args.force_ocr:
         kwargs["ocr_backend"] = "tesseract"
@@ -194,6 +199,7 @@ def main():
     parser.add_argument("--lang", default="eng", help="OCR language codes (default: eng)")
     parser.add_argument("--metadata", action="store_true", help="Include document metadata")
     parser.add_argument("--json", action="store_true", dest="as_json", help="Output as JSON")
+    parser.add_argument("--cache", action="store_true", help="Enable result caching to speed up repeated extractions")
     parser.add_argument("--check", action="store_true", help="Check installation and dependencies")
 
     args = parser.parse_args()

package/resources/skills/hd-docs-sync/SKILL.md

@@ -19,7 +19,7 @@ REQUEST → ASK → COLLECT → VERIFY → MAP → RECONCILE → APPLY → STAND
 | Phase              | Action                                  | Tools (Amp / Claude)                                          |
 | ------------------ | --------------------------------------- | ------------------------------------------------------------- |
 | 1. Ask             | Scope + start date from user            | (interactive)                                                 |
-| 2. Collect         | git log + history/ + plans/ (parallel)  | `Bash`, `Glob`, `Read`, `oracle` (Amp) / `Plan` subagent (Claude) |
+| 2. Collect         | git log + history/ + plans/ + related repos (parallel) | `Bash`, `Glob`, `Read`, `oracle` (Amp) / `Plan` subagent (Claude) |
 | 3. Verify          | Ground topics in code                   | `finder` (Amp) / `Explore` subagent (Claude)                  |
 | 4. Map             | Auto-detect target docs                 | `finder` (Amp) / `Explore` subagent (Claude)                  |
 | 5. Reconcile       | Topics vs code vs docs                  | `oracle` (Amp) / `Plan` subagent (Claude)                     |
@@ -77,9 +77,58 @@ Extract from output:
 2. Filter by date prefix >= start date
 3. `Read` any `.md` files found
 
+### Stream D — Related Repos Git History (conditional)
+
+**Check conditions first:**
+
+Read `additionalDirectories` from both `.claude/settings.json` and `.claude/settings.local.json`
+and merge into one flat list of paths.
+
+| AGENTS.md `## Related Projects` | `additionalDirectories` (merged) | Action |
+|----------------------------------|----------------------------------|--------|
+| ✅ present | ✅ has matching paths | Run git log for each matched path |
+| ✅ present | ❌ no matches found | **Warn user** (see below), skip Stream D |
+| ❌ missing | — | Silent skip |
+
+**Alias matching**: for each alias in `## Related Projects`, check if any path in
+`additionalDirectories` contains the alias as a directory name component
+(e.g., alias `project-b-api` matches `/some/path/project-b-api`).
+
+**Warning when Related Projects declared but no matching paths found:**
+
+```
+⚠️  Related Projects found in AGENTS.md but no matching paths in additionalDirectories.
+    Aliases declared: project-b-api, project-b-story-android
+    Stream D skipped — related repos git history will NOT be checked.
+
+    To fix: add to .claude/settings.json (committed, shared paths):
+    { "additionalDirectories": ["/path/to/project-b-api"] }
+
+    Or add to .claude/settings.local.json (gitignored, local paths):
+    { "additionalDirectories": ["/my/local/path/to/project-b-api"] }
+
+    See: /hd-help → Project Configuration Conventions → Multi-Repo Projects
+```
+
+Display this warning once, then continue with Streams A/B/C only.
+
+**When matches are found**, for each matched path:
+
+```bash
+git -C <resolved-path> log --since="<date>" --oneline --no-merges
+```
+
+Extract commit topics from the related repo. Label each with `[related: <alias>]`
+so downstream phases know the origin.
+
+**Why**: API changes in a related repo (e.g., renamed endpoint, changed request shape)
+can make the current repo's docs stale even when the current repo had no commits.
+
 ### Synthesis
 
-`oracle` (Amp) / `Plan` subagent (Claude) clusters all three streams into a unified topic list.
+`oracle` (Amp) / `Plan` subagent (Claude) clusters all four streams into a unified topic list.
+Stream D topics carry `[related: <alias>]` labels. Reconcile phase should note when
+a topic originates from a related repo rather than the current repo.
 See `reference/collection-prompts.md` → **Stream Synthesis Prompt**.
 
 ## Phase 3: Verify Against Code
@@ -125,6 +174,7 @@ Use file-change signals from Stream A to auto-target:
 | Troubleshooting insight | `docs/TROUBLESHOOTING.md` | — |
 | Auth / PII / payment / compliance changes | `SECURITY_STANDARDS.md` (suggest review only) | — |
 | Code style / patterns / convention changes | `docs/CODING_STANDARDS.md` — suggest review only | — |
+| Stack change (new language/framework added) | `docs/REVIEW_STANDARDS.md` — suggest updating `tech_stack` | — |
 
 > **CLAUDE.md**: This file is a short pointer to `AGENTS.md` and rarely needs updating. Only touch it if the project structure changes significantly (e.g., AGENTS.md is moved or renamed).
 >
@@ -175,7 +225,7 @@ For multiple unrelated files: spawn parallel `Task` per file.
 ## Phase 7: Standards Delta Check
 
 Runs at end of every `hd-docs-sync` execution — after Phase 6, independent of topic detection.
-Checks both standards pairs. If no delta on either: silent.
+Checks all three standards pairs. If no delta on any: silent.
 
 ### Standards pairs
 
@@ -183,6 +233,8 @@ Checks both standards pairs. If no delta on either: silent.
 | ---------------- | -------------- |
 | `skills/hd-security-review/SECURITY_STANDARDS.md` | `docs/SECURITY_STANDARDS.md` |
 | `skills/hd-code-review/CODING_STANDARDS.md` | `docs/CODING_STANDARDS.md` |
+| `skills/hd-code-review/REVIEW_STANDARDS.md` | `docs/REVIEW_STANDARDS.md` |
+| `docs/KNOWN_ISSUES.md` (format template from `hd-docs-init`) | `docs/KNOWN_ISSUES.md` |
 
 For each pair, run Sub-flow A or B:
 
@@ -200,6 +252,8 @@ On **yes**: `Read` base file → extract the project override scaffold (schema s
 
 On **no**: skip silently.
 
+For `docs/KNOWN_ISSUES.md` specifically: scaffold the file with the KI entry format template + an empty KI entries section (no example entries).
+
 ### Sub-flow B: Project file exists → semantic diff + suggest
 
 1. `Read` base file → extract semantic units (section rule titles + first sentence)
@@ -220,12 +274,19 @@ On approval: `Read` project file → `Edit` to append confirmed points under a
 
 Manual edit always allowed — developer may prefer to copy-paste directly.
 
+For `docs/KNOWN_ISSUES.md` specifically (Sub-flow B):
+- Parse all `## KI-NNN` headings and extract their `Target fix` dates.
+- Flag entries where `Target fix` date has passed → suggest resolving or updating the date.
+- Flag entries that still contain example placeholder text (e.g., "Example —") → remind dev to replace with real issues.
+
 ### Semantic parse rules
 
 | Standards file | Parse units as |
 | -------------- | -------------- |
 | `SECURITY_STANDARDS.md` | Section 3.x rule headings + Section 4 CRITICAL gate table rows |
 | `CODING_STANDARDS.md` | Section 2.x bullet-point groups (2.2, 2.3) + Section 3.x policy names |
+| `REVIEW_STANDARDS.md` | `tech_stack:` field + `reference/stacks/` preset names (Sub-flow A: offer create with auto-detected `tech_stack` pre-filled) |
+| `KNOWN_ISSUES.md` | `## KI-NNN` headings + `Target fix` date field per entry |
 
 Match criterion: project file contains the **section heading** (e.g. `3.9` or `Logging`) →
 consider that rule "present". No deep content comparison needed.
@@ -301,6 +362,7 @@ User: "Sync docs for the estimation enhancements from last 2 weeks"
 - [ ] Terminology matches existing docs
 - [ ] Code-is-truth rule applied (planned ≠ implemented)
 - [ ] Standards delta check completed (create or diff+suggest for both pairs)
+- [ ] Related repos git history checked (Stream D — if Related Projects + additionalDirectories configured)
 ```
 
 ## Troubleshooting

package/resources/skills/hd-docs-sync/reference/doc-mapping.md

@@ -15,6 +15,7 @@
 | Migration       | `docs/MIGRATION.md`                       |
 | Skills/workflow | Relevant `claude/skills/*/SKILL.md`       |
 | Release notes   | `CHANGELOG.md`                            |
+| Review standards | `docs/REVIEW_STANDARDS.md`              |
 
 ## By Change Type
 

package/resources/skills/hd-help/SKILL.md

@@ -58,7 +58,9 @@ Classify the user's description into one of these categories:
 | **QA** | test cases, QA, QC, tester, test suite, generate test cases, revise test cases, danger zones, verify test coverage, cross-reference tests, test case template |
 | **DailyGoals** | daily goals, morning goals, write today's goals, goals report, morning report before work |
 | **DailyReport** | daily report, daily standup, what I worked on, end of day report, done today, working on, blockers |
+| **DailyViewer** | view daily report, show past goals, yesterday's report, this week summary, daily history, review last week, what did I do, summary of the week |
 | **DocParse** | extract text, parse document, convert PDF to text, OCR document, read DOCX, parse spreadsheet, extract from PDF, convert to markdown, batch extract, kreuzberg |
+| **ISO** | ISO 9001, ISO 27001, compliance gap, iso lookup, iso onboard, iso change, sync iso docs, import google docs, iso policy, iso clause, hd-iso, iso verify, is this document ISO compliant, verify process against ISO, ncr, nonconformity, corrective action, mandatory records, certification readiness, are we ISO ready, audit readiness |
 
 ### Step 2: Route to Skill
 
@@ -98,6 +100,14 @@ Writing the morning goals report (what you plan to do today)?
 Writing the daily status report (what you did / are doing / blocked on)?
   YES → hd-daily-report
 
+Want to VIEW or SUMMARIZE past daily goals or reports?
+  YES → hd-daily-viewer
+    /hd-daily-viewer                  — list last 7 days with availability indicators
+    /hd-daily-viewer yesterday        — single day: goals + report + metrics
+    /hd-daily-viewer this week        — weekly summary with per-day breakdown
+    /hd-daily-viewer last 30 days     — monthly aggregates + outstanding activities
+    /hd-daily-viewer 2026-03-01       — specific date
+
 Something is broken, wrong, or failing?
   YES → hd-issue-resolution
 
@@ -149,6 +159,17 @@ Documentation goal:
   AGENTS.md exists but stale?     → hd-docs-sync
   Completed beads to log?         → hd-changelog
 
+ISO 9001 / 27001 compliance goal?
+  Sync Google Docs → Markdown?         → hd-iso-sync <folder-url>
+  Search ISO docs?                     → hd-iso lookup <query>
+  Check compliance gaps?               → hd-iso gap --standard 27001
+  Planning a doc change?               → hd-iso change <description>
+  Onboard a new employee?              → hd-iso onboard --role <role>
+  Verify doc content quality?          → hd-iso-verify <path|--clause|--standard|NL>
+  Log / track a nonconformity?         → hd-iso ncr (log | list | close | show)
+  Check mandatory ISO records exist?   → hd-iso records --standard 27001
+  Pre-audit readiness check?           → hd-iso-ready --standard 27001
+
 General problem exploration, unclear intent, or "I don't know what I need"?
   → hd-brainstorming  (good starting point for any ambiguous situation)
 ```
@@ -184,6 +205,7 @@ Use this table to populate the `**Requires:**` line above. Omit the line for ski
 | `hd-asana` | Python + `ASANA_API_KEY` env var |
 | `hd-look-at`, `hd-painter` | Python + `GEMINI_API_KEY` env var |
 | `hd-task`, `hd-tasks` | Linear / Jira / ClickUp / Asana MCP configured |
+| `hd-iso-sync` | Google Docs MCP (`a-bonus/google-docs-mcp`) + Google OAuth credentials |
 
 When requirements apply, append to the output:
 > **Requires:** `<from table above>` — see `docs/INSTALL.md` for setup
@@ -212,6 +234,8 @@ Use these rules when the routing step above has ambiguity:
 | `hd-look-at` vs `hd-painter` | Analyzing existing media (image/audio/video/screenshot)? → hd-look-at. Creating new images or videos from a text prompt? → hd-painter. |
 | `hd-docs-parse` vs `hd-look-at` | Extracting text content from a document (PDF, DOCX, XLSX, images via OCR)? → hd-docs-parse. Analyzing/describing visual or audio content (what's in this image, transcribe audio)? → hd-look-at. |
 | `hd-daily-goals` vs `hd-daily-report` | Morning, planning what to DO? → hd-daily-goals. During/end of day, reporting what was DONE? → hd-daily-report. |
+| `hd-daily-report` vs `hd-daily-viewer` | Writing today's report? → hd-daily-report. Viewing/reviewing a past report? → hd-daily-viewer. |
+| `hd-daily-goals` vs `hd-daily-viewer` | Writing today's goals? → hd-daily-goals. Reviewing past goals or summaries? → hd-daily-viewer. |
 
 ---